AI Audio, Voice Cloning, and You
Evan McGee @ SignalWire
This talk will engage the audience as we explore the current state-of-the-art when it comes to AI voices. voice cloning, and the associated technologies. We’ll start by examining the current OSS efforts and where they stand relative to the big commercial efforts from companies like ElevenLabs, Play.HT, Google, and others. We’ll then discuss the actual underlying AI technology powering these systems, including Mel-to-Mel Spectrographic training, style transfer, using vision-based AI to recognize errors, and more. The audience should leave with a great understanding of where/why/how to apply these to their various use cases and, with any luck, an optimism about the future of audio!
Evan McGee is CEO of Everlit and CTO @ SignalWire. Everlit is dedicated to bringing the written word to audio, working news, magazines, content creators, and more. SignalWire a company bringing programmable Voice, Messaging, and WebRTC communications to developers of all skill levels through the power of simple APIs.
Evan has over a decade of deep experience in the audio + telecommunications industry, including technical and business expertise with machine learning, AI, wireless MNO/MVNOs, traditional PSTN, and VoIP/WebRTC OSS. He has patents and patent applications covering both hardware and software designs and frequently presents at conferences on topics like containerized global SaaS deployments, microservices architectures, and the application of machine learning in realtime settings.